I wish I had something more profound about the bushfire situation.
I have put the sequence of MSLP/1000-500 thickness charts up on my website. It seems the fires kicked off on Christmas eve and during Christmas day in the northwesterly flow across the top of the low that was passing across south of Tassie.
MSLP charts: 01122300
01122400
01122412
01122500
01122512
01122600
As Milton pointed out, it is all related to the longwave trough sitting in the eastern bight. On Christmas day the Subtropical jet lay over South-eastern OZ, with a maximum over eastern NSW; and the jet-core and strong winds extended right down to 850 hPa and below.
TLAPS
200 Hpa 01122512
TLAPS
850 Hpa 01122512
What the underlying dynamics are, or even what the scientific question s are, I can only speculate on at this stage. It seems to get such a strong northwesterly gradient across southern NSW required substantial cold air advection to the west; and you can see that in the MSLP/thickness chart for the 12Z on the 24th. You can also see it in the surface theta chart for 25th 11 Z, which should be compared with the same chart for a few days earlier (22th 11z)
Theta Charts
01122511
01122211
An interesting aspect of
it is that as the low/front system moved in, we had a sort of a double-front
effect, with a lagging front down across the coral sea from the previous
system and a new front moving in with the Tassie-low. This
can be seen on the following
two charts for Christmas day:
Christmas
day charts: (i) large scale 850
anal showing a "frontal shear
line" from the Qld coast
across the Tasman and a new trough-front moving in
across southern NSW.
(ii) 900 hPa vorticity
Showing two fronts in the
region as blue lines of
cyclonic vorticity.
An interesting thing is that now, one week later the pattern has more or less repeated itself. The long-wave trough and the jet core are still in position, and as yesterday's front went through, we once again had the double frontal structure.
Chart: 900
vorticity: 011231
If anyone has any interesting perspectives on the situation, please let us know. It seems the dynamics are not those of a normal summer front, which has the parent low a long way further south. Rather the dynamics are tied in to the development of such a strong long-wave trough and unseasonally low-latitude subtropical jetstream core over southeastern Australia.
Did anyone write a report on the 94 fires? If so, could they send us the reference. Otherwise, with the nice NMOC chart archive that exists on the web now, just send us the key dates and we’ll have a look.
John McB
From Milton Speer
John,
There's the following reference:
Speer, Leslie, Colquhoun
and Mitchell, 1996:
The Sydney Australia Wildfires
of January 1994 - Meteorological Conditions and
High Resolution Numerical
Modeling Experiments. Int. J. Wildland Fire 6(3),
145-154.
M
From John McBride
I was talking to Graham Mills about the fire situation. He pointed out how dry it was in those westerlies ahead of the change. Looking at the observations for Camden posted earlier by Robin Hicks, the relative humidities were down below 10% for a while on Christmas day. The observations quoted by Elly for Sydney had dewpoints at -3C. Looking at the thetae map (see website), you can see an enormous dry tongue leaning across NSW up towards the Kimberley. What causes that: How did the air get so dry?
As Chris Webb pointed out the Sydney sounding on Christmas day showed dry adiabatic up to 640 hPa; so possibly much of the process is vertical mixing, such that the surface moisture gets diluted by being mixed through such a deep layer.
So... from Mills.... Maybe the scientific question we have to ask is: How did the lower layers of the atmosphere dry out to such a large extent?
Milton, any clues from your study of the 94 situation?
John McB
P.S.
How do you generate skew T - log P diagrams for past situations, e.g. Sydney
airport on Christmas day and Christmas Eve?
Date:
Wed, 02 Jan 2002 10:16:13 +1100
From:
Andrew Treloar <a.treloar@BoM.GOV.AU>
To:
Synoptic Discussion <synoptic_discussion@bom.gov.au>
Subject:
[synoptic_discussion] MesoLaps Wind Change for NSW this evening
Hi Everyone,
We are grappling with the possible development of a southerly wind change on the NSW coast this afternoon. The MesoLaps develops fresh/stronng southerlies during the afternoon, reaching the Illawarra during the evening. However, the MesoLaps doesn't pick up the strong burst of southerlies that went through Montague Island around 7am this morning.
Is the model picking up the later developments even though it has missed out a bit on initial developments? Observed winds over the NSW ranges are up to 50 kts at 700Hpa this morning which is consistent with the model, which also eases the upper winds over the Illawarra into the evening.
Any thoughts out there.
Andrew
--
Manager, NSW Weather Services
From Noel Davidson
G'Day Synopticians,
The
double-front structure that JMB mentioned is interesting.
The
comments below come out of a short discussion and some wild speculations
(based
on a cursory glance at a limited number of weather maps - not good practice,
but
I'll continue anyway) between G. Mills and myself this morning (2 Jan).
The
slowness of LAPS in forecasting the southerly change
reported
by Andrew T and Russell S, may also be related to this double front /
secondary
development issue.
1.
Those of us who have tried to understand and correct failed NWP forecasts,
know
what an horrendous task it can be. Problem areas in initial
conditions
are frequently non-local, making the search for the causes very difficult.
There
is also the uncertainty of inadequate representation of topographic
and
physical processes - which are almost impossible to deal with. [But
that
shouldn't stop us from having a go].
2.
There is some possibility that non-local analysis issues may have occurred
here.
It
appears that the first change may have weakened but accelerated in response
to the
development
of the upstream ridge and further upstream trough (the double
front
structure). If this is true then the poor forecast of the southerly change
may
be related to the far field rather than the local analysis. To forecast
this kind of
behaviour
would require forecasting the amplification of the upstream trough
and
the subsequent behaviour of the upstream ridge. So the structure problems
may have occurred around
45S,
130E (give or take 20 degrees!!). Such processes can develop and propagate
very
rapidly. Forecasting the behaviour then becomes related to accurate analysis,
not only
of
the system itself but also of systems over these poorly-observed areas.
3.
Forecasting the relative amplification of two "interacting" systems is
no easy task,
particularly
near complex topography. Indeed I'm sure most realise (but I'll say it
anyway)
that forecasting any developing, non-steady situation is difficult.
4.
Late news : Graham also reports that the very consistent (incorrect) forecast
behaviour
from
mesolaps perhaps suggests that topographic effects may not have been adequately
represented.
----------------------------------------------------------------------------------
I'm
not sure at this stage whether these comments are obvious and the speculations
wrong.
Of
course critical comments are welcomed.
One other issue has also arisen out of this discussion :
We
should have easy access to obs data, 6 hourly analyses and graphics
for
at least the previous 10 days - but we don't. So these near real-time
diagnoses and
discussions
become difficult, and presumably important forecaster feedback doesn't
happen.
Perhaps
we should encourage JMB to accelerate his establishing
comprehensive
data sets on his website that we could all easily view and discuss.
Indeed
perhaps this should be done using the various configurations of LAPS in
a more general
fashion.
Best
wishes
ND
(&GAM)
We understand that little time is available to forecasters to respond in any detail.
I believe
Graham will ask Kamal about LAPS ensemble for this case. Do we know
if
the observed southerly change fell within the envelope of the GASP ensembles?
